Interruptions during medication administration: a descriptive study.
The following databases were used: CINAHL Complete, Healthsource: Nursing/Academic Edition, Academic Search Complete, and Medline. Search terms included interruptions , medication administration, patient safety, and efficiency. The search was confined to studies published 2009-2016.
Biron, Lavoie-Tremblay, and Loiselle (2009) studied 102 medication administration rounds using direct observation at a tertiary care university teaching hospital in Quebec, Canada. They found nurses were interrupted at least once during 53.9% of observations, with an average of 6.3 work interruptions per hour. Nurse colleagues (29.3%) and system failures (22.8%; including missing medication or equipment) represented the most frequent interruptions during medication preparation. Nurses passing the medication (16.9%) and patients (16%) accounted for the most interruptions during medication administration. This study was noteworthy for pinpointing the actual source of the interruptions and documenting their frequency. However, it did not analyze interruptions by time of day or attempt to correlate number of frequencies and time for administration.
Nguyen, Connolly, and Wong (2010) studied 100 medication administration observations during a medication pass timeout at a 600 bed academic teaching hospital in California. Compared to 53.9% interruptions in the study by Biron and colleagues (2009), which had no timeout, 19% of the observations in this study had interruptions . Nguyen and co-authors (2010) also found reducing interruptions resulted in 100% of medications administered without errors. This study was important in demonstrating a timeout intervention could lead to a marked decrease in interruptions . However, the study did not discuss in detail the types of interruptions or note some may be necessary (e.g., colleague question needing immediate response).
Westbrook and co-authors (2010) used an observational technique with a sample of 4,271 medication passes at two major teaching hospitals in Sydney, Australia. They found the risk of a major clinical error during a single drug administration more than doubled, from 2.3% with no interruptions to 4.7% with four interruptions . However, Hopkinson and Jennings (2013) analyzed the earlier study and argued for a more nuanced examination of the relationship between interruptions and medication errors. The study by Westbrook and colleagues (2010) showed a general trend of increased errors with increased interruptions . Interestingly, researchers found clinical errors were lower with one or two interruptions than with no interruptions .
Kalisch and Aebersold (2010) observed nurses for 136 hours on seven units at two Midwestern hospitals. They found 10 interruptions per hour during all nurse task performance, not just during medication administration. Authors actually counted the number of errors per hour (1.5 in the first hospital, 1.89 in the second hospital). They also found significantly more interruptions (1 per 4.5 minutes/12 per hour in the first hospital, 1 per 13.3 minutes/4 per hour in the second hospital). No significant relationship existed between interruptions and medication errors, but researchers cautioned the complex multitasking environment increased potential for errors. The conclusion interruptions led to increased errors was not supported in this study.
At the time of this study, none of the research to date focused on the time of day of the medication administration. In addition, none of the studies included a statistical analysis of the relationship between interruptions and time for medication administration. The current study addressed both areas.
The purpose of this study was to determine the number and type of interruptions , and identify their effect on the time needed for medication administration. The study also analyzed the number and type of interruptions by time of day. Finally, the study was to include a statistical analysis of the relationship between interruptions and amount of time needed for medication administration.
This study employed direct observation of medication administration. Direct observation can produce valid, reliable results (Feleke, Mulatu, & Yesmaw, 2015; McLeod, Barber, & Franklin, 2015).
The study site was a 271-bed acute care Magnet[R] hospital. Nurses cared for an average of six patients. Medical-surgical units were targeted because of concerns about their efficiency of medication passes.
Medication administration was observed for 30 patients on two telemetry (40%, n=12), one medical (36.6%, n=11), and one combined medical-surgical-oncology unit (23.3%, n=7). In addition, 52.6% (n=20) were distributed at 9:00 a.m.; 3.3% (n=1) at 1:00 p.m.; 13.3% (n=4) at 5:00 or 6:00 p.m.; 13.3% (n=4) at 9:00 p.m.; and 3.3% (n=l) at 1:00 a.m. Times and distribution of observations were consistent with the percentage of medications administered in a 24-hour period.
The study was presented to the Institutional Review Board (IRB) at Presence Saint Francis Hospital (Evanston, IL) with the checklist for data collection. The IRB approved the study; members determined no need existed for nursing consent because there was no deviation from normal practice, data on nursing identity were not collected, and the standard of care for medication administration did not change. No patient information was collected.
A tool developed by the California Nursing Outcomes Coalition (Raban & Westbrook, 2013) was used for this study, with the following indicators:
* Comparison of medication to the medication administration record (MAR) before administration
* Lack of distraction or interruption during medication administration
* Nurse checking of two forms of patient identification
* Explanation of medication to the patient
* Charting of medication on the MAR immediately after administration
The representative workflow modeled after the California Nursing Outcomes Coalition tool was reviewed with a nursing administrator and several senior clinical nurses from the study site, chosen because they knew the medication workflow on the study units. The checklist based on this model was evaluated for validity by nurse staff and the Director of Nursing.
Data collectors were educated concerning the workflow and purpose of the study. Each point on the tool was identified. Minor changes to the tool included the addition of times at every point. The checklist (see Table 1) was evaluated for reliability by having data collectors observe identical medication passes and complete the tool. Inter-rater reliability was 100% with 10 observations.
Nurses administering medications at identified times on the designated units were chosen for observation. Nurses starting their medication pass when the data collector arrived were chosen for observation. Nurses were observed assessing laboratory results, vital signs, and patient physical symptoms; obtaining and preparing the medication; comparing the medication to the electronically generated MAR; administering the medication; and documenting the administration manually. All parts of the observations were timed.
Frequency of Interruptions
No interruptions occurred in 37% (n= 11) of the 30 medication passes. One or two interruptions occurred in 33% (n=10) of the passes; three to five interruptions occurred in 30% (n=9). Mean interruptions per medication pass were 1.50 (5D=1.61).
Correlation of Number of Interruptions with Time to Administer Medications
The medication pass was timed from when the medication was obtained until it was documented on the MAR. Routine medication passes could include multiple medications. Mean time from medication retrieval to documentation was 15.5 minutes, with one observation of medication pass taking 70 minutes (SD= 15.1, n=24). This calculation did not include the entire 30pass sample size because data on documentation were not available for six medication passes. Median time from retrieval to documentation was 11 minutes. See Figure 1 for median times from obtaining the medication to documentation; median times were used because of the effect of outliers on a small sample. A statistically significant correlation between the number of interruptions and the time from obtaining the medication to documentation (Spearman's rho=0.404, p=0.05, n=24), as well as a statistically significant correlation for the time from obtaining medication to administration (Spearman's rho=0.404, p=0.05, n=30), were found. The sample size was smaller for the former group because data on documentation were not available for six medication passes.
Interruptions by Shift and Unit Type
Interruptions were more frequent during the day shift than the evening shift (see Figure 2). Day shift interruptions were from telephone calls, other staff, physicians, families, and other procedures. Evening-night shift interruptions were from other staff and telephone calls. Nurses on the day shift also were interrupted by interprofessional staff seeking information or performing patient care. Telemetry units had fewer instances of more than two interruptions than either medical-surgical units or medical-surgical oncology units (see Figure 3).
Interruptions to Retrieve Supplies
Nurses retrieved supplies not on the cart during 43% of the observations (n=30). When supplies were obtained, the mean time from obtaining the medication to medication administration increased from 5.7 (SD=5.6) to 10.6 minutes (SD=5.4); median time increased from 4 to 11 minutes. Nurse planning influenced the need to interrupt the medication pass to obtain supplies. Some nurses stocked their carts before beginning their medication passes; others continually returned during passes to the nurses' station for supplies (e.g., water, applesauce, syringes).
Interruptions to Perform Additional Assessment and Interventions
All nurses spent time reviewing each medication with the electronic printout of the MAR. Nurses also monitored orders for new medications that should be included in the current pass. Some situations called for additional assessment to hold medications or request changes in an order if patients experienced adverse effects or nurses learned of potential contraindications.
Discussion of Findings
Interruptions were widespread during medication passes on four medical-surgical units in a community hospital, similar to findings of Smeulers and colleagues (2013) and Verweij, Smeulers, Maaskant, and Vermeulen (2014). Sources of interruptions included talking directly with physicians, family members, or other staff; taking telephone calls from staff, family, or nursing homes; obtaining supplies; obtaining patients' assessment information; stopping to fulfill patient requests; and seeking clarification of orders. By comparison, recent studies found primary sources of interruptions to be other nurses (Hall et al., 2010; McLeod et al., 2015) and the patients themselves (McLeod et al., 2015). Other interruption sources identified by researchers included transport personnel and alarms (Federwisch, Ramos, & Adams, 2014).
[FIGURE 2 OMITTED]
The type of interruption should be analyzed carefully because interruptions may cause a negative or positive patient outcome (Hopkinson & Jennings, 2013). Beneficial interruptions include being informed of laboratory results that would contraindicate giving a medication or dose or being provided information of a patient's earlier symptom (e.g., diarrhea) that could warrant withholding the medication. In fact, communication during a medication pass may be the only time for critical conversations between physicians and nurses. Positive interruptions also may come from machines (e.g., cardiac monitors, intravenous pumps) notifying the nurse of updated patient information that would warrant a delay in medication administration (Jennings, Sandelowski, & Mark, 2011). Eliminating these interruptions could lead to negative patient outcomes (Freeman et al., 2013).
Clinical nurses, managers, and directors reviewed and prioritized distractions experienced during medication administration to eliminate interruptions that could have a negative effect on patient outcomes. An action plan was developed that included screening of telephone calls during prime medication administration times. Calls were categorized as follows:
* Emergent calls continued to go directly to the nurse (e.g., physician, critical laboratory results, patient-family calls).
* Calls patient care technicians and unit secretaries could address were routed to them (e.g., transportation).
* Other calls were delayed for the nurse to return later (e.g., nursing home follow up, personal calls). To minimize unnecessary interruptions , nurses received patient reports from transferring units, ensured medications were available, and collected necessary assessment parameters (vital signs and pertinent laboratory values) before beginning the routine medication pass.
Medication administration is a complex process for medical-surgical nurses caring for an average of six patients each. During the task, nurses also analyze adverse medication effects, assess patients' understanding of the medication and its side effects, and meet knowledge deficits in preparation for safe administration by the patient when discharged.
The following additional recommendations may help minimize unnecessary interruptions :
* Nurses should stock carts with supplies before beginning medication rounds.
* Consideration should be made for the availability of nursing assistants during the 9:00 a.m. medication pass. In this study, because this medication pass coincided with morning care, nurses often were required to assume tasks nursing assistants otherwise would manage.
* Messages to nurses during medication administration can be in the form of less obtrusive texting rather than telephone calls (Tomietto, Sartor, Mazzocoli, & Palese, 2012).
* Nurses should know the schedule of patients leaving the unit for physical therapy, x-ray, or procedures to give them their medications first. These events thus would not become an interruption.
Nurses were aware they were being observed. The direct observation technique has been criticized for altering the behavior of the observed subject. However, the effect can be negligible if observers are experienced, objective, unobtrusive, and non judgmental (Westbrook et al., 2010). The small sample size was another limitation. A larger number of observed nurses could have contributed to more robust findings.
Recommendations for Future Research
More research is needed to determine if recommended changes to reduce interruptions will lead to increased efficiency and reduced medication errors. More research also is needed on the relationship between interruptions during medication administration and patient outcomes. Such research could analyze the type of interruption and determine if it impeded or facilitated the nurse's role in providing patient care.
Nurses must be flexible and available to meet ever-changing patient conditions (Federwisch et al., 2014). Communicating with physicians, reporting critical laboratory values, receiving important messages from family members, or learning about substantial patient concerns from other nursing staff may be necessary despite their occurrence during medication administration. However, interruptions should be delayed if they are not essential to current medication administration or patient outcome.
The purpose of this study was to quantify medication administration interruptions on four medical-surgical units and examine their effect on efficiency in the medication administration process.
Interruptions decrease the efficiency of the medication process workflow. Research on the relationship between interruptions and medication errors is not conclusive. Therefore, nurses examining the role of interruptions in safety, as opposed to efficiency, need to be mindful of this distinction.
Researchers used direct observation of nurses obtaining, passing, and documenting the administration of patient medications on four medical-surgical units (two telemetry units, one medical-surgical unit, and one combined medical-surgical/oncology unit) in a 271-bed Magnet[R] community hospital in suburban Chicago.
Interruptions occurred in 63% of the medication passes (n=30). The mean number of interruptions per medication pass was 1.50 (SD=1.61, CI 0.8975-2.1025). The median number of minutes from obtaining the medication to documentation was 7 minutes, increasing to 10.5 minutes with one to two interruptions , and to 16 minutes with three to five interruptions . A positive correlation between interruptions and medication pass time was found; thus, the more times the nurse was interrupted, the longer the time required for the medication pass (Spearman's rho=0.404, p=0.05; n=30).
In this study, interruptions during medication passes resulted in an increased amount of time to pass medications, decreasing efficiency. An action plan was developed that focused on increased preparation before beginning medication rounds along with reduction of unnecessary interruptions during medication rounds. Necessary interruptions , such as calls from physicians, critical lab results, and calls from families, still were permitted. This separation of necessary from unnecessary interruptions was based on research that questions the view all interruptions have negative consequences.
Biron, A.D., Lavoie-Tremblay, M., & Loiselle, C.G. (2009). Characteristics of work interruptions during medication administration. Journal of Nursing Scholarship, 41(4), 330-336.
Clark, A.P., & Flanders, S. (2012). Interruptions and medication errors: Part II. Clinical Nurse Specialist, 26(5), 239-243. doi:10.1097/NUR.0b013e31825e5be4 Colligan, L., & Bass, E.J. (2012). Interruption handling strategies during paediatric medication administration. British Medical Journal Quality and Safety, 21(11), 912917.
Federwisch, M., Ramos, H., & Adams, S.C. (2014) . The sterile cockpit: An effective approach to reducing medication errors? American Journal of Nursing, 114(2), 4755.
Feleke, S.A., Mulatu, M.A., & Yesmaw, Y.S. (2015) . Medication administration error: Magnitude and associated factors among nurses in Ethiopia. BMC Nursing, 14(53), 1-8. doi: 10.1186/s12912-015-0099-1
Fore, A.M., Sculli, G.L., Albee, D., & Neily, J. (2013). Improving patient safety using the sterile cockpit principle during medication administration: A collaborative, unit-based project. Journal of Nursing Management, 21(1), 106-111.
Freeman, R., McKee, S., Lee-Lehner, B., & Pesenecker, J. (2013). Reducing interruptions to improve medication safety. Journal of Nursing Care Quality, 28(2), 176-185.
Flail, L., Ferguson-Pare, M., Peter, E., White, D., Besner, J., Chisholm, A. Plemingway, A. (2010). Going blank: Factors contributing to interruptions to nurses' work and related outcomes. Journal of Nursinq Management, 18(8), 1040-1047.
Plopkinson, S.G., & Jennings, B.M. (2013). Interruptions during nurses' work: A state-of-the-science review. Research in Nursing & Health, 36(1), 38-53.
Jennings, B.M., Sandelowski, M., & Mark, B. (2011). The nurse's medication day. Qualitative Health Research, 21 (10), 1441-1451.
Kalisch, B.J., & Aebersold, M. (2010). Interruptions and multitasking in nursing care. The Joint Commission Journal on Quality and Patient Safety, 36(3), 126132.
McComas, J., Riingen, M., & Kim, S.C. (2014). Impact of an electronic medication administration record on medication administration efficiency and errors. Computers Informatics Nursing, 32(12), 589-595.
McLeod, M., Barber, N., & Franklin, B.D. (2015). Facilitators and barriers to safe medication administration to hospital inpatients: A mixed methods study of nurses' medication administration processes and systems (the MAPS study). PLoS ONE, 10(6), e0128958. doi:10.1371/journal. pone.0128958
Nguyen, E.E., Connolly, P.M., & Wong, V. (2010). Medication safety initiative in reducing medication errors. Journal of Nursing Care Quality, 25(3), 224-230.
Raban, M.Z., & Westbrook, J.I. (2013). Are interventions to reduce interruptions and errors during medication administration effective? A systematic review. BMJ Quality & Safety, 23(5), 414-421. doi:10.1136/bmjqs-2013-002118
Smeulers, M., Hoekstra, M., van Dijk, E., Overkamp, F., & Vermeulen, H. (2013). Interruptions during hospital nurses' medication administration rounds. Nursing Reports, 3(e4), 18-23.
Tomietto, M., Sartor, A., Mazzocoli, E., & Palese, A. (2012). Paradoxical effects of a hospital-based, multi-intervention programme aimed at reducing medication round interruptions . Journal of Nursing Management, 20(3), 335-343.
Verweij, L, Smeulers, M., Maaskant, J.M., & Vermeulen, FI., (2014). Quiet please! Drug round tabards: Are they effective and accepted? A mixed method study. Journal of Nursing Scholarship, 46(5), 340-348.
Westbrook, J.I., Woods, A., Rob, M. I., Dunsmuir, W.T.M., & Day, R.O. (2010). Association of interruptions with an increased risk and severity of medication administration errors. Archives of Internal Medicine, 170(8), 683-690.
Christine Hughes Cooper, MBA, MSN, RN-BC, is Lead Specialist, Patient Care Informatics, Presence Healthcare, DesPlaines, IL.
Ruth Tupper, MS RN, is Research and Education Nurse, Presence Saint Francis Hospital, Evanston, IL.
Karyn Holm, PhD, RN FAAN, FAHA, is Professor Emeritus, School of Nursing, DePaul University, Chicago, IL; and Research Consultant, Presence St. Francis Hospital, Evanston, IL.
TABLE 1. Data Collection Tool Date: RN reviews medications  yes  no on MAR RN obtains medication  med drawer  Pyxis RN compares medication  yes  no to MAR Supplies obtained  yes  supplies requested RN takes medication  med cart  tray to patient RN assesses patient  yes  no before administration 2nd RN assesses dose  yes  no RN identifies patient  checks name band  asks name RN reviews medications  yes  no with patient RN administers medication  yes  no Medication route  IV  topical  IVPB Frequency  stat  PRN RN records administration  yes  no on MAR RN charts in system  blood sugar  vitals Interruptions to RN  phone  physician Patient response recorded  yes  no Nursing unit Date: RN reviews medications on MAR RN obtains medication  refrigerator RN compares medication to MAR Supplies obtained RN takes medication  pocket to patient RN assesses patient before administration 2nd RN assesses dose  NA RN identifies patient  asks date of birth RN reviews medications with patient RN administers medication Medication route  oral Frequency  routine RN records administration on MAR RN charts in system  reason medication not given Interruptions to RN  patient family  another patient Patient response recorded Nursing unit Date: Time RN reviews medications on MAR RN obtains medication  bedside RN compares medication to MAR Supplies obtained RN takes medication to patient RN assesses patient before administration 2nd RN assesses dose RN identifies patient  other RN reviews medications with patient RN administers medication Medication route  other Frequency  single dose RN records administration on MAR RN charts in system Interruptions to RN  another RN/staff Patient response recorded Nursing unit FIGURE 1 Median Minutes from Obtaining the Medication to documentation by Number of Interruptions n=7 7 1-2=10 10.5 3-5 n=7 16 Note: Table made from bar graph. FIGURE 3. Percentage of Medication Passes Interrupted by Unit Type Telemetry Medical Medical-Surgical, Oncology n=12 n=11 n=7 0 42 36 28.5 1-2 50 18 28.5 2+ 8 46 43 Note: Table made from bar graph.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Research for Practice|
|Author:||Cooper, Christine Hughes; Tupper, Ruth; Holm, Karyn|
|Date:||May 1, 2016|
|Previous Article:||Digestive enzyme replacement therapy: pancreatic enzymes and lactase.|
|Next Article:||The problem with heart failure.|